Maurer, Matthew J.

Background: The goal of clinical trial eligibility criteria is to identify a population of patients that have high risk disease or have unmet needs with standard of care therapy. Frontline trials of diffuse large B-cell lymphoma (DLBCL) include patients with International Prognostic Index (IPI) scores of 3-5 as a way to identify a high risk population. Trials typically include patients with IPI score=2, and sometimes IPI score=1, to expand the eligible patient population. However, broader patient populations must be balanced by the potential inclusion of low-risk patients with excellent outcomes to standard of care therapy, leading to potentially underpowered trials. IPI is a sum of dichotomous variables which makes implementation easy but can result in heterogeneity in outcomes within IPI scores. This heterogeneity may be caused by unequal prognostic information across IPI variables, loss of information within IPI variables due to the use of cutoffs, as well as information captured by variables that are not included in the IPI. In prior modeling of outcomes in DLBCL (Maurer et al, AJH 2016), we identified that: i) LDH has a strong prognostic ability beyond the upper limit of normal (ULN) and ii) bulky disease is an independent variable for prognosis. Here we evaluate if the presence of either of these two variables, easily captured at diagnosis within the current standard clinical work-up, can identify a high-risk subset of patients within the population of those with IPI scores of 1-2. Methods: Patients with newly diagnosed lymphoma were enrolled within 6 months of diagnosis in the LEO Cohort at 8 academic medical centers from 2015-2020 and prospectively followed. Inclusion criteria for this analysis was aggressive B-cell lymphoma typically included in frontline trials (e.g. DLBCL, high grade B-cell lymphoma, FL3B), IPI 1-5, frontline therapy with rituximab and anthracycline-based chemotherapy, stage II-IV disease and ages 18-80 years. Event-free survival (EFS) was defined as the time from diagnosis until progression, initiation of 2 nd line therapy, or death due to any cause; EFS was evaluated at 24 months (EFS24). Very high LDH was defined as LDH>1.3xULN based on prior modeling (Maurer et al, AJH 2016) and bulky disease was defined as maximum tumor diameter (MTD) of 7cm or greater. IPI 1 and 2 were first evaluated separately and then combined due to similar outcomes in patients with very high LDH and/or bulky disease. High risk IPI1-2 was defined as the presence of either bulky disease and/or very high LDH within an IPI score of 1-2. Confirmation was performed in a similar cohort of patients enrolled in the Iowa/Mayo SPORE Molecular Epidemiology Resource (MER) from 2002-2015; bulky disease was defined using a 10cm cutoff per availability in MER data collection forms. Results: 1560 patients from LEO were analyzed; median age was 64 years (IQR: 54-71), 42% were female and 83% had stage III/IV disease; subtype was DLBCL, NOS in 1335 (86%). IPI was 1-2 in 766 patients, 3 in 496 patients and 4-5 in 298 patients ( table). 391 (51%) with IPI1-2 had either bulky disease and/or very high LDH (high risk IPI1-2) and 375 had low-risk IPI1-2. EFS24 in high risk IPI1-2 (71.6%, 95% CI: 67.2-76.3) was significantly inferior to IPI1-2 without bulky disease or very high LDH (85.2, 95% CI: 81.7-88.9 logrank p=0.0035, figure) and similar to patients with IPI3 (70.7%, 95% CI: 66.8-74.9). Consistent results were observed in 1014 patients analyzed in the MER using a higher cutoff for bulky disease (10cm) due to data availability ( table). Patients with high risk IPI1-2 had significantly shorter diagnosis to treatment interval (DTI) in both the LEO (per-week OR=0.83, 95% CI: 0.77-0.88) and MER (per-week OR=0.77, 95% CI: 0.68-0.87) cohorts compared to low risk IPI1-2, both p<0.0001). In sensitivity analysis, LEO results were similar when performed in the subset of patients with DLBCL, NOS. Conclusion: The presence of either bulky disease and/or very high LDH at presentation identifies a high-risk subset of patients with IPI 1-2 that have similar outcomes to patients with IPI 3 disease. These simple criteria can be easily implemented within existing clinical trial diagnostic work-ups. Future trial designs in frontline aggressive B-cell lymphoma should include patients with IPI 3-5 and patients with IPI 1-2 with either bulky disease and/or very high LDH to better define a high-risk population of patients for clinical trial eligibility.

Purpose: Improved predictive tools are desired in classic Hodgkin lymphoma (cHL) to improve prognostication and to better risk stratify patients (pts) for clinical trials. In younger pts who tolerate intensive therapies, focus has been on limiting late treatment toxicity through risk-adapted, PET-guided treatment strategy. In older pts, acute treatment toxicities are limiting factors for optimal disease control. Due to these fundamental differences, predictive tools may perform differently in younger and older pts. In particular, tools developed in younger pts may be suboptimal in older populations and vice versa. The primary objective of this study was to assess the performance of the A-HIPI (development was limited to 18-65 years of age) against the international prognostic score (IPS) for modelling overall survival (OS) and progression-free survival (PFS) in younger and older pts with cHL. Methods: This study utilized data from the population-based, nationwide Danish National Lymphoma registry. Pts with advanced-stage cHL (Ann Arbor stage III-IV or IIB with extra nodal disease) were included and categorized by age in two groups: ages ≤65 years vs. >65 years. In the primary analysis, all pts regardless of therapy were included. A sensitivity analysis restricting to pts starting on BEACOPP or ABVD was performed (Only ABVD in the >65 cohort). Concordance index (C-index) and calibration were used to assess model performance and generalizability. Calibration was visualized using calibration plots and quantified by calibration in the large (CITL) and calibration slope (CS) at 5-years post diagnosis. Results: A total of 893 advanced-stage HL pts fulfilled the inclusion criteria (634 in the ≤65 cohort and 259 in the >65 cohort). The 5-year OS and PFS in the >65-year cohort were 44% (Confidence intervals (CI); 37-51%) and 35% (CI; 30-42%), respectively. Fifty-six pts in the >65 cohort were classified as low-intermediate-risk (IPS; 1-2), 128 intermediate-risk (IPS; 3-4), and 75 high-risk (IPS; 5-7). For the ≤65 cohort, the 5-year OS and PFS was 88% (CI; 85-90%) and 76% (CI; 73-80%), respectively. Twenty-three pts were classified as low-risk (IPS; 0), 244 low-intermediate-risk, 286 intermediate-risk, and 81 high-risk. The A-HIPI had a C-index of 0.74 and 0.63 for OS and PFS, respectively in the ≤65 years cohort vs. 0.65 and 0.64 for OS and PFS, respectively, in the >65 cohort. The IPS had a C-index in the ≤65 cohort of 0.66 and 0.60 for OS and PFS, respectively, vs. 0.58 and 0.59 for OS and PFS, respectively, in the >65 cohort (i.e., the IPS performed better in the ≤65 cohort in terms of OS and PFS but was inferior to the A-HIPI). When restricting to pts treated with ABVD or BEACOPP, the A-HIPI C-index in the >65 cohort (116 pts) were reduced to 0.55 for OS and 0.54 for PFS. For the ≤65 cohort (595 pts), the A-HIPI C-index was stable at 0.75 and 0.62 for OS and PFS, respectively. In terms of A-HIPI calibration in the >65 cohort, CITL and CS were -0.78 and 0.39 for OS vs. -0.98 and 0.80 for PFS. In the ≤65 cohort, CITL and CS were -0.06 and 0.99 for OS vs. 0.21 and 1.18 for PFS. Visual inspection of the calibration plot (Figure 1) for the >65 cohort showed that the A-HIPI underestimated the observed proportion for both OS and PFS (i.e., the observed events were higher than the predicted). Calibration plot for the ≤65 cohort (Figure 2) indicated that the A-HIPI accurately modelled the observed proportion for both OS and PFS (i.e., the predicted probability was consistent with the observed proportion). Conclusion: The newly developed A-HIPI was validated in a large population-based cohort of Danish pts with advanced-stage cHL. Performance of the A-HIPI was high in the ≤65 cohort and superior to the IPS for OS prediction. Calibration of A-HIPI was excellent in pts ages ≤65 years. However, A-HIPI performed significantly worse among pts >65 years and attained poor performance in pts receiving ABVD or BEACOPP, which is not surprising since patients older than 65 were not included in the development. This study emphasizes the importance of further efforts to develop or update models that apply specifically in older individuals. The inclusion of additional variables reflecting comorbidities and pts fitness into a prognostic model may be necessary to effectively deliver accurate prognoses and risk assessments for the older pt.

Background: Understanding the symptoms and quality of life (QOL) over time of adults with follicular lymphoma (FL) is important for treatment decision-making and clinical management. However, there are limited population-level data on the long-term QOL of adults with FL. We aim to describe the real-world, long-term QOL of adults with newly diagnosed FL up to 5/6 years after diagnosis. Methods: We used the Mayo Clinic/Iowa Molecular Epidemiology Resource (MER) and the multi-institutional Lymphoma Epidemiology of Outcomes (LEO) to identify adults with grade 1-3A FL who completed QOL surveys at baseline and follow-up. Participants with lymphoma were prospectively enrolled within 9 months of diagnosis in the MER cohort from 2002-2015 and within 6 months of diagnosis in the multi-institutional LEO cohort study across 8 cancer centers from 2015-2020. The LEO/MER cohort studies systematically collected information on disease status, QOL, health behaviors, and functional assessment. Treatments, disease relapses, and deaths were verified by medical record review. Treating physicians determined clinical management. QOL was measured using the Functional Assessment of Cancer Therapy-General (FACT-G) at years 1, 2, 3, and 5/6 post-diagnosis. A higher FACT-G total score (range 0-108) indicated better QOL across 4 subscales (range): physical (0-28), social/family (0-28), emotional (0-24), and functional (0-28) well-being. We categorized participants based on their frontline management at time of diagnosis: “observation” (surveillance), “treatment” (systemic treatment with immunotherapy +/- chemotherapy), and “local” (radiation) groups. We employed a generalized linear mixed model to evaluate and compare the changes in QOL scores (as a continuous variable) from baseline for the observation, treatment, and local groups, adjusting for sex, race/ethnicity, age, FLIPI risk, and cohort (i.e., LEO or MER). We also evaluated QOL changes by FLIPI risk. Results: Our study included 1,544 participants with FL and QOL data. At the time of enrollment, median age was 61 years (range 19-91), 88% were non-Hispanic White, and 49% were female. Based on initial management, 529 (34%) were in the observation group, 880 (57%) were in the treatment group, and 135 (9%) were in the local group. FLIPI risk was high for 17% in the observation group, 29.3% in the treatment group, and 3% in the local group. The mean (standard deviation) baseline FACT-G total score was lowest in the treatment group with the following baseline scores: 86 (13) for the observation group, 83 (14) for the treatment group, and 88 (12) for the local group (Table 1). The observation and local groups reported a greater worsening of their baseline FACT-G total score vs the treatment group at both the 2-year (-1.2, -0.8 vs +1.8, respectively) and 5/6 year (-5.4, -6 vs. -1.6, respectively) timepoints (Figure 1). This appeared to be driven by worsened physical and functional well-being in the observation (-1.3, -2.2, respectively) and local (-0.9, -2.4, respectively) groups vs the treatment group (0, -0.3, respectively). These differences were statistically significant between the observation vs treatment groups. There were no statistically significant differences in the other subscales, i.e., social/family and emotional well-being, by frontline management. Social/family well-being decreased across all groups by 5/6 years (-2.3 in observation, -2.2 in treatment, and -3 in local, p<0.05). Regardless of FLIPI risk, the total FACT-G scores decreased within each FLIPI risk group at 5/6 years (-4.4 for low, -3.8 for intermediate, and -4.9 for high, p<0.05). This was primarily due to worsened social/family well-being (-2.4 for low, -2.2 for intermediate, and -3 for high, p<0.05) at 5/6 years. There was no statistically significant difference in QOL between FLIPI risk groups. Conclusion: This is one of the first and largest studies with real-world longitudinal QOL data for FL. Our study suggests that frontline systemic treatment initially improved QOL and resulted in a lower degree of QOL decline over time. Social/family well-being decreased over time regardless of frontline management or FLIPI risk. Further research is warranted to explore the impact of timing and specific systemic treatments on QOL, incorporate the patient/caregiver experience to address social/family well-being, and identify the clinical significance of our findings.

Background: To be useful across the broadest range of patients and clinical settings, clinical prediction models should be externally validated in different data sets and clinical settings and refined over time. In 2022, the HoLISTIC Consortium (www.hodgkinconsortium.org) developed and validated the A-HIPI model via TRIPOD guidelines for adults 18-65 with advanced stage classic Hodgkin Lymphoma (AS-HL) (Rodday et al., JCO 2023). The A-HIPI model, based on individual patient (pt) data from more than 5000 patients (pts) from North America, Europe, and Australia, predicts progression-free survival and overall survival within the first 5 years (PFS5 and OS5) based on pre-treatment pt and disease characteristics (www.qxmd.com/calculate/calculator_869/a-hipi). In collaboration with Brazilian clinical experts and researchers, we describe the performance of the A-HIPI in the Brazilian Prospective HL registry. Methods: The Brazilian Prospective HL registry (NCT02589548) was originally established in 2009 with the goal of evaluating risk factors, treatment, and disease outcomes. By 2022, 1357 pts with classic HL were registered with a median follow-up of 5 years for surviving patients (Biasoli I. HemaSphere 2022). 5-year PFS and OS were calculated using the Kaplan-Meier method. Using the methodology for validation and calibration outlined in by Rodday et al., calculated risk scores were calculated for PFS5 and OS5. Harrell c-statistic and the Uno c-statistic were calculated to evaluate discrimination (i.e., differentiation of those at higher risk of having an event from those at lower risk) and plots of deciles of predicted vs. observed outcomes were created to assess calibration (i.e., how similar the predicted absolute risk of the model is to the true (observed) risk in an external cohort). In the primary A-HIPI analysis, laboratory values outside of permitted range were excluded and not imputed. This was checked in sensitivity analysis using multiple imputation for missing values; no substantial differences were noted. Thus, in the current validation, out-of-range lab values were excluded. There were no exclusions in the current analysis for early death, defined as deaths that occurred from the onset of first treatment until 30 days after its end. Pts with HIV were excluded from the assembled cohort. Results: 694 patients with AS-HL with the inclusion criteria of the original study were included in the present study. The median age was 30.9 years (range, 18-65) at diagnosis with 46% of the pts being female. 230 (33%) patients were classified as Stage IIB, 190 with Stage III (27.4%), and 274 (39.5%) with Stage IV. Bulky disease (>10 cm) was present in 226 patients (33%). Median hemoglobin was 11.6 g/dL (8-16.3), median albumin 3.63 g/dL (1.41-5.46), and median lymphocyte count was 1.5 x 10 3 µL (0.10-4.76). The 5-y PFS (PFS5) in the A-HIPI Brazilian validation dataset was 68.4% (95% CI: 64.7-71.9%) and the 5-y OS (OS5) was 86.0% (95% CI: 82.8-88.6).Harrell c-statistics for the A-HIPI model were 0.60 (95% CI, .56-.64) for 5-year PFS and 0.69 (95%CI, .64-.74) for 5-year OS and Uno c-statistics for the A-HIPI model were 0.63 (95% CI, .58-.68) for 5-year PFS and 0.72 (95%CI, .65-.78) for 5-year OS. Calibration plots (see figures) show the predicted A-HIPI model estimates were well calibrated with the observed outcomes in the Brazilian HL registry. Conclusions: Despite important differences between the original development cohort for A-HIPI and the Brazilian registry in patient and clinical characteristics (e.g., younger age at diagnosis, greater proportion of Stage IV disease) and the well characterized impact of income maldistribution and socio-economic status on clinical outcomes (Biasoli et al., Int J Cancer 2018), the A-HIPI clinical prediction tool performed well in the Brazilian Prospective Registry. This included evidence of adequate discrimination and calibration. Future studies are planned, including the impact of treatment selection and response, use of interim imaging, and supportive care may underscore important differences in outcomes among adults with AS-HL.

Background: Relapsed DLBCL occurs in up to 40% of patients and has a poor prognosis. Current prognostic tools such as the international prognostic index (IPI) and newer forms (NCCN-IPI) only include baseline clinical features. PET/CT features have not yet been incorporated into prognostic risk models despite being essential for staging and response assessment. Deep learning computer algorithms have consistently demonstrated the ability to analyze imaging data and identify features that are not readily apparent to the human eye. Using a deep learning computer vision model that automatically extracted radiomic features from PET/CTs, we examined the association of changes between baseline PET and interim PET (PET2) radiomic features on survival outcomes and compared the prognostic performance with NCCN-IPI. Methods: Adult patients with newly diagnosed diffuse large B-cell lymphoma (DLBCL) between 2005 and 2015 and enrolled in the prospective MER cohort were included. Native DICOM images from FDG PET/CT scans at baseline and interim PET (PET2, after 2-3 cycles of chemotherapy) were preprocessed for regularity and then segmented using the nnU-Net architecture with an external lymphoma-specific model. Feature extraction and standard uptake value (SUV) characterization were accomplished with PyRadiomics and NiBabel. Our computational method included no manual segmentation correction. Individual radiomics features (n=115) and the % change above (Δ High) and below (Δ low) 50 th quartile from baseline to PET2 were evaluated for association with event free survival (EFS) and overall survival (OS) using a univariate Cox regression model and by Kaplan Meier analysis. EFS was defined as time from PET2 to first systemic relapse or death from any cause. OS was defined as time from PET2 to death from any cause. Additionally, Spearman and Pearson correlation matrices were used to guide feature selection while minimizing collinearity. We then assessed the association of the highest scoring radiomics features with EFS in a multivariate cox regression model. Results: A total of 506 patients had baseline and PET2 scans available for radiomic feature extraction. Median age at diagnosis was 61 years (range 18-90), 43% were females. NCCN-IPI composition and outcomes of our cohort was comparable to the previously published NCCN cohort, with patient distribution and 5-year OS as follows: Low (9%, 5-y OS 83%), Low-intermediate (38%, 5-y OS 49%), High-intermediate (41%, 5-y OS 31%), High (11%, 5-y OS 12%). At a median follow up of 81 months,there were 308 events, and272 patients had died.Several radiomics features were associated with EFS including change in tumor surface area (SA Δ) (HR: 1.75, 95% CI 1.36-2.24, p<0.005, median of 26,927 mm 2, Figure 1A) and change in total metabolic tumor volume (TMTV Δ) (HR: 1.64, 95% CI 1.27-2.10, p<0.005, median of 120,473 mm 3, Figure 1B). As a reference, NCCN-IPI univariate Cox regression analysis of EFS produced a Harrell's C-statistic (c) of 0.624. Single radiomics features of SA Δ and TMTV Δ individually produced unadjusted c-statistics of 0.616 and 0.614 respectively. Combinations of radiomics features (SA Δ, TMTV Δ and NCCN-IPI classification) demonstrated cumulative benefit (c=0.652). Conclusion: Automatic PET-CT radiomic feature extraction using deep learning models is feasible among patients with newly diagnosed DLBCL. The change in tumor SA and TMTV appear to be promising radiomic biomarkers. A low decline in either tumor SA or TMTV by PET2 is associated with higher risk of relapse and poor survival outcomes. As individual characteristics, these features are similar prognostic indicators of EFS as the NCCN-IPI. Further investigation into risk stratification when combining radiomics with clinical and genomic variables is warranted. Figure 1. Event free survival for DLBCL patients with A) a change in tumor surface area by PET2 less than 50% percentile (SA_Δ Low) and greater than 50 th percentile (SA_Δ High) and B) a change in TMTV by PET2 less than 50% percentile (TMTV_Δ Low) and greater than 50 th percentile (TMTV_Δ High).

Background: Event-free survival at 24 months (EFS24) after frontline immunochemotherapy (IC) is an important endpoint in diffuse large B-cell lymphoma and follicular lymphoma, and patients who achieve EFS24 have similar survival compared to age- and sex-matched general population. The role of EFS24 in predicting subsequent survival has not been established in mantle cell lymphoma (MCL), possibly due to perceived poor survival historically. In the last decade, the outcomes of MCL are improving in the evolving treatment landscape. In this international multicenter study, we investigated conditional survival and cause of death in patients with MCL who achieved EFS24 after frontline IC in the older and more recent eras. Methods: Outcomes after frontline IC were evaluated in 5 independent cohorts that included over 3000 patients in total: Mayo/Iowa MER prospective cohort, BC Cancer retrospective population-based cohort, US 12-center retrospective cohort, Swedish Lymphoma Registry, and Danish National Lymphoma Registry. For each cohort, 2 treatment eras were defined based on cohort-specific shifts in treatment patterns (Table 1). Overall survival (OS) after diagnosis and after achieving EFS24 were compared to the background age- and sex-matched general population using a standardized mortality ratio (SMR). Cumulative incidences of cause-specific deaths were analyzed using a competing risk model. Results: In the MER cohort, patients treated in Era M1 (2002-2009; n=147) and Era M2 (2010-2015; n=153) both had inferior OS compared to the general US population. A lower SMR in Era M2 vs M1 (2.25 vs 3.46) suggests a narrower gap in OS compared to the general US population. Lymphoma was the leading cause of death in both eras. In Era M1, patients who achieved EFS24 still had inferior OS compared to the general US population (SMR=2.40, 95% CI 1.76-3.19), and lymphoma remained the leading cause of death. For patients in Era M2 who achieved EFS24, the difference in OS compared to the general US population was not statistically significant with current follow-up (SMR=1.43, 95% CI 0.93-2.09), and lymphoma was no longer the leading cause of death (Table 2). In the BC cohort, patients treated in Era B2 (6/2013-2019; n=188) vs Era B1 (2003-5/2013; n=250) had a narrower gap in OS compared to the general British Columbia population (SMR 4.53 vs 6.69). Lymphoma was the leading cause of death in both eras. For patients achieving EFS24, the gap in OS was narrower in Era B2 vs B1 (SMR 3.56 vs 4.99). After achieving EFS24, lymphoma was the single leading cause of death for patients in Era B1 but not in Era B2 (Table 2). In the US 12-center cohort, patients treated in Era U1 (2002-2011; n=312) and Era U2 (2012-2016; n=417) both had inferior OS compared to the general US population (SMR 2.68 and 2.92, respectively). In patients who achieved EFS24, with current follow up, the difference in OS compared to the general US population was statistically significant in Era U1 (SMR=2.01, 95% CI 1.50-2.64) but not in Era U2 (SMR=1.44, 95% CI 0.82-2.34). Lymphoma was not the leading cause of death after achieving EFS24 (Table 2). In the Swedish cohort, the gap in OS compared to the general Swedish population was narrower in Era S2 (2013-2018; n=439) vs S1 (2006-2012; n=442), both after frontline IC (SMR 4.8 vs 5.4) and after achieving EFS24 (SMR 2.6 vs 3.4). After achieving EFS24, lymphoma was the leading cause of death for patients in Era S1 but not in Era S2 (Table 2). In the Danish cohort, patients treated in Era D2 (2014-2020; n=370) vs Era D1 (2004-2013; n=461) had a slightly narrower gap in OS compared to the general Danish population (SMR 1.90 vs 2.10). For patients who achieved EFS24, those in Era D1 still had inferior OS compared to the general Danish population (SMR=1.44, 95% CI 1.22-1.68), but the OS difference compared to the general Danish population in Era D2 was not statistically significant (SMR=1.27, 95% CI 0.92-1.72) with current follow-up (Table 2). Cause of death data were not available in this cohort. Conclusion: Survival in patients with MCL who achieved EFS24 after frontline IC improved in the more recent treatment era and moved closer to the background expected survival. After achieving EFS24, lymphoma-related mortality was no longer the leading cause of death in the more recent era. EFS24 following frontline treatment may become a critical endpoint for predicting subsequent outcomes in patients with MCL in the modern era.

Background: Shorter diagnosis to treatment interval (DTI) has been shown to be associated with more aggressive disease features and inferior outcomes in diffuse large B-cell lymphoma (DLBCL; Maurer et al, JCO 2018). Additional confirmatory studies have also shown that DTI is related to tumor volume (Alig et al, JCO 2021) and molecular phenotype (Aluaij et al, Blood 2023). Here we evaluate DTI in a large multicenter cohort of patients prospectively enrolled and followed as part of the Lymphoma Epidemiology of Outcome (LEO) cohort. Methods: Patients were enrolled within 6 months of diagnosis at one the 8 LEO cohort academic medical centers in the United States between 2015 and 2020. Patients in this analysis had aggressive B-cell lymphoma (BCL) treated with anthracycline-based chemotherapy. DTI was defined as the initial lymphoma biopsy date until the start of initial chemotherapy; patients with DTI between 0-100 days were evaluated. Short DTI was defined as DTI ≤ 14 days (DTI≤14) as previously described in prior publications (e.g. Maurer et al, JCO 2018) and confirmed via examination of functional forms via splines. Event-free survival (EFS) was defined as the time from start of treatment until progression/relapse, retreatment, or death due to any cause. Overall survival (OS) was defined as the time from start of treatment until death due to any cause. Results: 2565 patients with aggressive B-cell lymphoma were evaluated. Median age was 62 years (IQR 51-71) and 1469 (57%) were male. 297 patients (12%) were self-reported non-White race and 314 patients (12%) were self-reported Hispanic or Latino. The majority had DLBCL, NOS subtype (N=1927, 75%), while 227 (9%) had high-grade BCL with MYC and BCL2 and/or BCL6 rearrangements (HG, DH) and 90 (4%) had high-grade BCL NOS (HG, NOS). Clinical characteristics are summarized in the table. At median follow-up of 49 months (IQR 36-68), 800 patients (31%) had an event and 578 patients (23%) died. EFS at 24 months (EFS24) was 76% (95% CI: 74-77). Median DTI across all subtypes was 21 days (IQR 12-33) and 845 patients (33%) had DTI≤14. Patients with DTI≤14 were significantly younger (median age 59 years) but there was no association between DTI and gender, race, or ethnicity. Patients with DTI≤14 were more likely to have HG, DH or HG NOS subtypes, ECOG PS 2-4, advanced stage, elevated LDH, CNS involvement, B symptoms, and bulky disease ( see table). Patients with DTI≤14 were more likely to receive R-EPOCH-based or intensive therapies (41%) and less likely to receive 1L therapy on a clinical trial (3.8%) compared to DTI>14 (26% and 10.1%, respectively). Consistent with previously reported data, DTI≤14 was associated with inferior EFS (HR=1.62, 95% CI: 1.41-1.87, figure), OS (HR=1.73, 95% CI: 1.47-2.04), and EFS24 (OR=1.93, 95% CI: 1.60-2.38). The association between DTI and outcomes remained significant after stratifying for subtype (EFS HR=1.68, 95% CI: 1.46-1.94; OS HR=1.78, 95% CI: 1.50-2.10) and adjustment for IPI (EFS HR=1.43, 95% CI: 1.24-1.65; OS HR=1.47, 95% CI: 1.24-1.74). In subset analysis, DTI≤14 was significantly associated with inferior outcomes within the subtypes of HG, DH (EFS HR=2.70, 95% CI: 1.82-4.01; OS HR=2.97, 95% CI: 1.92-4.60); HG, NOS (EFS HR=2.07, 95% CI: 0.93-4.60; OS HR=2.36, 95% CI: 1.03-5.43); and DLBCL, NOS (EFS HR=1.51, 95% CI: 1.27-1.78; OS HR=1.52, 95% CI: 1.45-1.85). Analysis was limited in other subtypes due to number of events. Notably, within the subset of patients with DTI>14, there was no significant difference in EFS between subtypes for DLBCL, NOS (EFS24=80%, EFS HR=ref), HG, DH (EFS24=73%, EFS HR=1.26, 95% CI: 0.92-1.73) and HG, NOS (EFS24=83%, EFS HR =0.84, 95% CI: 0.40-1.78), logrank p=0.29; similar results were observed for OS (p=0.22). Conclusions: Patients requiring early initiation of therapy for aggressive B-cell lymphoma represent a distinct population of patients with more aggressive clinical features and inferior outcomes. In patients with longer DTI, high grade subtypes had similar outcomes to DLBCL, NOS. Efforts should be made to include patients with anticipated short DTI in clinical trials and translational studies to fully capture the spectrum of patients with aggressive B-cell lymphoma.

Background: There are a lack of clinical prediction tools in early-stage Hodgkin lymphoma (E-HL). We developed and validated a modern-day model, known as E-HIPI, to predict progression-free survival (PFS) within the first 5 years (y) incorporating detailed individual patient (pt) data from international clinical trials and prospective registry data that were standardized, normalized & harmonized as part of the HoLISTIC Consortium (www.hodgkinconsortium.org). Methods: Model development utilized Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) guidelines (Moons, Ann Intern Med 2015) on a cohort of 3,080 newly-diagnosed E-HL pts treated on 4 seminal phase 3 clinical trials (NCIC: Meyer, NEJM 2012; RAPID: Radford, NEJM 2015; EORTC-GELA H9U: Ferme, E ur J Cancer 2017; EORTC/LYSA/FIL H10: Andre, JCO 2017). External validation was done in an independent cohort of 462 E-HL pts from a large registry (Princess Margaret, Toronto). Pts with classic HL, stage I or II disease, and ages 18-65 y were included. The primary outcome was PFS. Cox models were utilized with follow-up truncated at 5 years. Baseline predictors were: sex, stage, B symptoms, histology, number & location of nodal sites, and continuous data values of age, maximum tumor diameter (MTD), white blood cell count (WBC), lymphocyte count, hemoglobin, albumin & erythrocyte sedimentation rate (ESR). Multiple imputation was used for missing data. To consider possible non-linear relationships for continuous variables we examined plots & cubic splines. We used backward elimination ( P<0.1) to develop the models & internal validation with 200 bootstrap samples was conducted to estimate optimism & correct for overfitting. The final prediction equations applied optimism corrections to beta coefficients & hazard ratios (HR). Sensitivity analyses included model results stratified by PET staging (vs CT) and each HL trial. The models were evaluated in the external validation cohort & compared using c-statistics. Model scores were also grouped by tertiles for visualization in Kaplan-Meier (KM) curves. Results: Median age in the development cohort was 35y; 51% were female; 81% had nodular sclerosis & 77% stage II disease; 26% had B symptoms; 32% had mediastinal bulk (≥10cm or >1/3 diameter) and the mean MTD was 6.4 cm (interquartile range (IQR), 3.9-8.2). Median follow-up was 60 months (IQR 45-75). KM estimates at 5y was 90.7% (264 events) for PFS, and 97.1% (78 events) for overall survival (OS). After backward elimination, significant variables retained were age, sex, nodal location (cervical region), MTD & albumin. Stage, ESR, or the number of nodal sites were not significant. Age was modeled with piecewise linear splines due to its non-linear relationship. See the Table for the final prediction model and optimism-corrected HRs. Optimism-corrected c-statistics in the development model was 0.63. Sensitivity analyses of stratification by type of imaging at staging or by each clinical trial demonstrated no change in model parameters or c-statistics. Most baseline characteristics and outcomes of the external validation cohort were similar besides median follow-up time (74 months, IQR 37-129). C-statistics for the E-HIPI model in external validation was 0.63. In addition, observed 5y outcomes stratified by tertile of predicted risk in the external cohort separated pts into a low & higher-risk group, which comprised 1/3 & 2/3 of pts, respectively ( Figure). Finally, model diagnostics revealed the potential for time dependency in the retained covariates. Further assessment of differential prediction between early (<2y) vs. later (2-5y) PFS events is ongoing. Conclusion: Following TRIPOD methodology, we rigorously developed and externally validated the E-HIPI among >3,500 E-HL pts.Discrete risk groups were delineated, and we identified several novel factors, including nodal location, sex & key continuous variables that predict 5y PFS. Additional model refinement is underway, including formal calibration & external validation via other HL registries (Stanford, Mayo/Iowa SPORE). A web-based calculator to simplify application of the E-HIPI will also be presented at ASH. Future planned modeling includes integration of differing treatments & response-adapted imaging via multistate modeling to enrich individualized survival estimates and simulation modeling to predict late consequences.

Background: Tumor mutational burden (TMB), defined as the number of nonsynonymous mutations per megabase (Mb) in the coding regions of a tumor genome, has been associated with improved response to immune checkpoint blockade treatments and more broadly with overall survival (OS) in many cancers. Limited prior studies have found that TMB is higher in diffuse large B-cell lymphoma (DLBCL) compared to many other cancers. However, only two studies have evaluated TMB and prognosis in newly diagnosed DLBCL patients. The prognostic role of TMB in diffuse large B-cell lymphoma (DLBCL) remains unclear as prior studies have been limited to panel sequencing of cancer genes and/or lacked matched germline-tumor samples for filtering. Methods: We characterized TMB from whole exome sequencing (WES) in newly diagnosed DLBCL patients with matched diagnostic tumor-normal samples prior to treatment with standard immunochemotherapy (IC). Tumor DNA was extracted from formalin-fixed, paraffin embedded tissue sections and germline DNA from peripheral blood. We evaluated the association between TMB and event-free (EFS) and OS using Cox regression adjusted for the international prognostic index (IPI) and other factors including cell of origin (COO), MYC double hit (DH) with BLC2 or BCL6 by FISH, and TP53 mutation status. EFS and OS were defined as time from diagnosis to earliest of treatment failure (EFS only), death, or last follow-up; EFS at 24 months (EFS24) was also modeled as a marker of early aggressive DLBCL. An optimal high/low TMB cutpoint was determined based on EFS using maximally selected rank statistics; sensitivity analyses used an 20/80 percentile to define the threshold. We also evaluated the association of driver tumor mutational load (driverTML) with EFS and OS, defined as the total number of mutations in 292 DLBCL driver genes. Results: In the analysis cohort of 394 DLBCL patients, the median age at diagnosis was 65 years (interquartile range: 56-73), 43% were female, 62.2% had IPI 0-2, and 57.4% germinal center B-cell COO. MYC was DH in 7.5% and TP53 was mutated in 16.2% of patients. Through a median follow-up of 7.5 years, there were 141 events and 122 deaths. EFS24 was achieved in 75.1% of the cohort; 5-year EFS was 69% (95% CI 64-74%) and 5-year OS was 75% (95% CI 71-80%). Median TMB in our cohort of DLBCL patients was 3.6 mutations/Mb, which was the fifth highest compared with 32 cancers in The Cancer Genome Atlas. As a continuous variable, increasing TMB (mutation/Mb) was associated with superior EFS (HR=0.89, 95% CI 0.81-0.98) and OS (HR=0.89, 95% CI 0.81-0.99) after adjustment for IPI. Spline analyses confirmed linear associations of TMB with EFS and OS. The optimal modeled cutpoint based on EFS was 4.5 mutations/Mb, with 31% classified as high TMB (>4.5). No significant ( P >.05) differences were observed between high and low TMB patients with clinical (i.e., age, sex, IPI category) or molecular (cell of origin, MYC DH, TP53 mutated) factors. In Cox regression, high-TMB was significantly protective for EFS and for OS compared to low-TMB after adjustment for each clinical factor ( Figure 1), with simultaneous adjustment for all these factors associated with superior EFS (HR=0.52, 95% CI 0.33-0.84) and OS (HR=0.59, 95% CI 0.36-1.00). TMB (high/low) combined with IPI (0-2, 3-5) was significantly ( P =.001) associated with EFS in Kaplan-Meier analysis ( Figure 2). Among patients with IPI 3-5, high TMB was significantly associated with superior EFS (HR=0.47, 95% CI 0.27-0.81) and OS (HR=0.57, 95% CI 0.33-0.99). Among patients with IPI 0-2, we also observed longer EFS (HR=0.62, 95% CI 0.34-1.11) and OS (HR=0.62, 95% CI 0.32-1.20) in high TMB patients, although the associations were not statistically significant ( P >.05). In sensitivity analyses, a 20/80 percentile TMB threshold found similar results for EFS (HR=0.64, 95% CI 0.41-1.03) and OS (HR=0.69, 95% CI 0.42-1.12) as in our main analyses. TMB and driverTML were significantly correlated ( R=0.65, P <10 -16) and we found similar effect sizes for EFS (HR=0.45, 95% CI 0.22-0.91) and OS (HR=0.55, 95% 0.27-1.13) as in our main analyses. Conclusions: In a large and well-defined cohort of IC treated DLBCL patients, we found DLBCL patients with higher TMB had superior EFS and OS after adjustment for other clinical and molecular prognostic factors, and the prognostic association for TMB was robust to threshold and gene set.

Background: Primary central nervous system lymphoma (PCNSL) is a rare form of non-Hodgkin lymphoma with two-year overall survival (OS) rates of 66-70% reported in prospective clinical trials. Induction therapy with high-dose methotrexate (HD-MTX) followed by consolidation has become the mainstay of treatment, however, there are few contemporary and robust prognostication indices widely available for clinical use. Some of the most used indices were developed prior to the routine use of rituximab and include the International Extranodal Lymphoma Study Group (IELSG; 1980-1999), Memorial Sloan-Kettering Cancer Center (MSKCC; 1983-2003), Nottingham/Barcelona (NB; 1986-2001), and Taipei Score (TS; 2003-2015). Herein, we describe a contemporary comparison of clinical prognostication indices using a single-center cohort of PCNSL patients treated at the Mayo Clinic. Methods: Patients with a diagnosis of PCNSL who received HD-MTX as part of induction therapy at Mayo Clinic between October 2010 and June 2022 were identified and retrospectively reviewed. Patients with prior or concurrent systemic lymphoma were excluded. Primary endpoints were progression-free survival (PFS) defined as time from diagnosis to relapse, progression, or death due to any cause; and OS defined as time from diagnosis to death due to any cause. PFS and OS were evaluated using Kaplan-Meier curves and compared by risk scores using a log-rank test. The association between baseline characteristics and survival outcome were assessed in univariate Cox regression models. Any characteristics showing significant (p-value <0.05) association with outcome were combined using a multivariate Cox regression model. All characteristics that maintained significant association with outcome were combined into a new prognostic model. We compared the performance of each prognostic index model using Harrell's C (HC). Results: A total of 148 patients were identified (50.7% female) with median age at diagnosis of 66 years (range 29-85), with 48 patients (32.4%) age >70 (Table 1). Most had multifocal disease (60.1%) and deep brain involvement (67.6%) at diagnosis with elevated CSF protein (58.1%) but no documented CSF involvement (93.9%). In total, 117 patients (79.1%) received methotrexate, rituximab, and temozolomide (MRT) induction therapy, and the remaining 31 patients (20.9%) received methotrexate and rituximab (MR) induction therapy with various consolidation strategies. The median follow-up for all patients was 4.5 years and the two-year OS was 79% (95% CI: 73-86%). This cohort was stratified using previously published prognostication indices (Table 1). Upon analysis, PFS was not accurately predicted using the IELSG (score 4-5: HR 0.7 (CI 0.3-2.0); HC 0.6), MSKCC (3: 1.3 (0.4-4.0); 0.5), NB (3: 0.9 (0.4-2.5); 0.6), or TS (3: 0.9 (0.1-7.2); 0.5) prognostication indices. Furthermore, OS was not accurately predicted using the IELSG (4-5: 1.5 (0.6-4.0); 0.6), MSKCC (3: 1.2 (0.5-3.2); 0.6), NB (3: 1.2 (0.5-3.2); 0.6), or TS (3: 1.3 (0.2-9.8); 0.6) prognostication indices. Upon multivariate analysis, no variables listed in Table 1 were associated with PFS within this cohort. The two variables predicting a worsened OS were an ECOG performance status of 2 or more (1.9 (1.01-3.6)) and an MR induction regimen instead of MRT (2.1 (1.05-4.3)) (Figure 1). The two-year OS with no risk factors was 86.1% (95% CI: 78.8-94.1%), one risk factor was 75.7% (95% CI: 63.6-90.1%), and two risk factors was 61% (95% CI: 43-86%) (Figure 1). Conclusions: This study reports overall improved outcomes in patients with PCNSL compared to prior prospective and retrospective cohorts, likely in part due to treatment advances and supportive care strategies. This single-center, retrospective cohort of rituximab-exposed patients with PCNSL were not well prognosticated using previously published indices, indicating that further study and more sophisticated approaches to prognostication are warranted. Although multivariate analysis did not demonstrate any variables associated with PFS, it did demonstrate an inferior OS with a performance score of 2 or more and in patients treated with MR rather than MRT.